Joined Flow’s early engineering team (<10 → 70+). Work across multiple teams and functions designing foundational systems for Flow’s digital twin of real-world agreements that serves as the programmable layer for workflows across Flow’s digital twin of buildings, assets, and diverse lines of business (not just multifamily).

Conceived a declarative DSL for operating on Flow’s entity graph and contributed to the design of AI-native ontologies and primitives for agreements and other core entities.

Led technical design and validation strategy for complex migrations (involving async + event-driven data pipelines), and built reconciliation and verification tooling to ensure correctness of financial general ledger and contractual data.

Core interviewer for Flow’s code review loop, evaluating software design and engineering judgment. Conducted 40+ interviews and trained additional interviewers.

https://a16z.com/announcement/flow/

Tech Lead for human-in-the-loop review and label generation products at Abnormal:

· Defining the roadmap and building the infrastructure + tooling that supports a world-class human feedback loop, resulting in superior threat detection efficacy for Abnormal’s customers and constant improvement of our ML systems over time

· Invested in growing new engineers at the company through both one-on-one and structured learning initiatives that I proposed and led

Used: Python (Django, Celery), JavaScript (React), Redis, SQL, Databricks, Terraform, Kubernetes, AWS

(now defunct) Some folks had a “web3 gap year”; I had a browser gap year 🫡

First engineer hire for Chromium-based multiplayer browser. Owned changes to Chromium, devised and wrote related tooling for building and rebasing, designed and built new chrome.XX APIs to support multiplayer features, built desktop app update infrastructure, filmed demoes, did user research, helped hire engineers 2-4.

Key contributor to user-facing products, such as Looker's Slack integration 💜 (https://cloud.google.com/blog/products/data-analytics/slack-integration):

· Wrote APIs for mapping custom Slack commands to Looker queries, frontend and backend for its App Home, link unfurling, linking multiple Slack workspace accounts to Looker accounts and installations.

· Added timing and retry metrics for calls to external services in a multi-tenant proxy service used for routing all Slack events to a Looker instance, and wrote the first draft of the service runbook.

· Added product analytics for the Slack integration and wrote the main document for describing user interaction types in our usage analytics store.

· Fixed bugs in Looker's viz render and delivery (scheduled dashboards, S3, email, webhook, etc.) pipeline.

Used: Kotlin (ktor), Ruby, TypeScript (React), Slack APIs (OAuth flow, Block Kit), Datadog